Workflow Analysis - Inner Shift

Screen 01

Frame

Define who is in the room, what workflow is being analyzed, and what outcome matters.

Privacy: Do not enter unnecessary PHI, PII, confidential identifiers, account numbers, or other sensitive record-level data.

Client / Interviewee *

Organization *

Role / Title *

Industry / Function *

Consultant *

Consultation Date *

I confirm I will not enter unnecessary sensitive or personally identifying information. *

Workflow Name *What single repeatable workflow are we examining?

Trigger *What causes this process to begin?

Completion / End State *What tells you this process is complete?

Person Served *Who ultimately receives the value?

Intended Outcome *If this works exceptionally well, what should be true at the end?

Failure Condition *What happens when the workflow performs poorly?

Frequency

Approx. Staff Time per Cycle

Screen 02

Map

Build the real current-state workflow from a recent example.

Use a real case: “Walk me through the last time this happened. What happened first? What happened next?”

Screen 03

Diagnose

Capture friction evidence and score each material workflow step.

Friction Evidence

Mark each category Present, Not Present, or Unknown. If Present, capture evidence.

Step Scores

Six 1–5 scores are required for every material step.

Screen 04

Bound

Determine what context, authority, capability, and human presence must remain protected.

Human Judgment Reserve Statement *Complete: AI may support ______, but a human must retain authority over ______.

Screen 05

Decide

Finalize step dispositions, review generated recommendations, add consultant recommendations, and define success.

Step Dispositions

System-Surfaced Recommendations

The widget surfaces up to three. Consultant judgment remains final.

Consultant Recommendations

Add up to three additional recommendations using the same evidence standard.

Success Metrics

At least one measurable outcome is required.

Screen 06

Act

Test implementation reality, determine the right level of intervention, surface evidence gaps, and finalize.

Current Systems *

Systems That Must Remain

Sensitive / Regulated Information? *

Type of Sensitive Information

Known AI Policy / Restriction? *

Policy / Restriction Detail

Required Approvers

Primary Implementation Barrier *

Commercial Disposition Scores

Value1 low benefit · 5 significant value

Workflow Clarity1 poorly understood · 5 clear

Implementation Complexity1 simple · 5 cross-functional/complex

AI Specificity1 no clear use case · 5 specific/testable

Human Reserve / ConsequenceDerived from workflow; consultant confirms

Scope1 one workflow · 5 enterprise pattern

Governance Triggers Count

30–90 Day Success Measurable?

Bounded AI Use Case?

Consultant Final Commercial Disposition *

Secondary / Subsequent Disposition

Override RationaleRequired if consultant changes the system result.

Client-Facing Rationale *

Evidence Gaps

Completion Dashboard

Finalize

Save as Preliminary Analysis if additional evidence is still required before implementation.